Horizontally transferred cell-free chromatin particles function as autonomous ‘satellite genomes’ and vehicles for transposable elements within host cells
- Translational Research Laboratory Advanced Centre for Treatment, Research and Education in Cancer Tata Memorial Centre Kharghar, India
- Homi Bhabha National Institute, Anushakti Nagar, India
Figures
Figure 1
Electron microscopy image of cfChPs isolated from pooled serum of patients with cancer.
A ‘beads-on-a-string’ appearance typical of chromatin is clearly seen. Reproduced with permission from Mittra et al., 2015b.
© 2015, Indraneel Mittra. Figure 1 was reprinted with permission from Supplementary Figure 2B in Mittra et al., 2015b, Journal of Biosciences, which was published under a CC-BY-NC-ND. Further reproductions must adhere to the terms of this license.
Figure 2
Abundant uptake of cfChPs by NIH3T3 cells at 6 hr.
The DNA and histones of cfChPs were dually fluorescently labelled with Platinum Bright 550 and ATTO-488, respectively, and applied (10 ng) to NIH3T3 cells. (a) Representative images of dually labelled cfChPs prior to their application to NIH3T3 cells. (b) Confocal microscopy images of cfChPs-treated cells at 6 hr showing many dually labelled fluorescent signals in the cytoplasm and nuclei. (c) Fluorescence microscopy images of chromatin fibres prepared from similarly treated cells at 6 hr showing numerous dually labelled fluorescent signals of varying sizes in the cytoplasm and in association with the chromatin fibres. (d) Fluorescence microscopy images of metaphase spreads prepared from cfChPs-treated cells at 6 hr showing multiple dually labelled fluorescent signals, which are either associated with the chromosomes or are present in extrachromosomal spaces. The latter are marked with arrows.
Figure 3 with 3 supplements
Internalised cfChPs combine to form complex concatemers.
(a) FISH analysis of chromatin fibres prepared from cfChPs-treated NIH3T3 cells in continuous passage using different combinations of human chromosome-specific FISH probes, including probes specific for the human centromere and telomere, revealing co-localised fluorescent signals (arrows). (b) Small cfChPs also show co-localised signals suggesting that they too are comprised of concatemers (c) Similar co-localising signals are shown on metaphase spreads. Co-localised signals are marked with arrows.
Figure 3—figure supplement 1
Chromosome-specific probes that were used are human specific.
Chromatin fibres were prepared from HEK-293 (human embryonic kidney) and NIH3T3 (mouse fibroblast) cells and both cell types were probed with different pairs of human chromosome specific FISH probes and those against human telomere and centromere. (a) Positive control cells (HEK-293); (b) Negative control cells (NIH3T3).
Figure 3—figure supplement 2
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of different pairs of chromosomes.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 3—figure supplement 3
Control experiments to detect human DNA signals in control NIH3T3 cells that had not been exposed to cfChPs.
The human-specific FISH probes did not detect any human DNA signals in mouse cells.
Figure 4 with 1 supplement
Variable spatial relationships of concatemers with mouse chromatin fibres.
(a) Immuno-FISH analysis of chromatin fibres prepared from cfChPs-treated NIH3T3 cells in continuous passages using an antibody against histone H4 and a human whole-genome FISH probe. (b) Immuno-FISH analysis of chromatin fibres prepared from clone D5 showing that the concatemers persisted even after numerous passages and freeze–thaw cycles.
Figure 4—figure supplement 1
Control experiments to show that the genomic DNA FISH probe used was human specific and did not cross-react with mouse.
(a) Positive control experiment using HEK-293 (human embryonic kidney) cells treated with human specific DNA FISH probe; (b) Negative control experiment using NIH3T3 (mouse embryonic fibroblast) cells treated with mouse specific DNA FISH probe.
Figure 5 with 3 supplements
The concatemers synthesise DNA and express human DNA polymerase γ.
(a) Chromatin fibres were prepared from NIH3T3 cells in continuous passage and were pulse-labelled for 24 hr with BrdU (10 µm). Immuno-FISH experiments using antibodies against BrdU, human-specific DNA polymerase γ, and a human whole-genome FISH probe show co-localised signals of human DNA and BrdU, human DNA and DNA polymerase γ, and BrdU and DNA polymerase γ. Co-localised signals are marked with arrows. (b) Similar co-localising signals are shown on metaphase spreads. Co-localised signals are marked with arrows.
Figure 5—figure supplement 1
Control experiments to show that the antibody used against DNA Polymerase γ was human specific and did not cross-react with mouse.
(a) Positive control experiment using HEK-293 (human embryonic kidney) cells probed with human specific DNA FISH probe, antibody against BrdU and antibody against human specific DNA Polymerase γ; (b) Negative control experiment using NIH3T3 (mouse embryonic fibroblast) cells do not react with either human specific DNA FISH probe or with human specific DNA Polymerase γ antibody.
Figure 5—figure supplement 2
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of human DNA and BrdU, human DNA and DNA polymerase, and BrdU and DNA polymerase.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 5—figure supplement 3
Control experiments to detect human DNA signals in control NIH3T3 cells that had not been exposed to cfChPs.
The human-specific FISH probe and antibody did not detect any human DNA and DNA polymerase signals in mouse cells.
Figure 6
Concatemers proliferate and amplify themselves within the mouse genome over time.
Metaphase spreads were prepared from cells in each progressively increasing passage and probed with a human-specific DNA FISH probe. (a) The total number of human DNA FISH signals on the DAPI-stained chromosomes was counted, and the mean number of human FISH signals per chromosome was calculated after analysing 10 metaphase spreads at each passage. A progressive increase in mean FISH signals per chromosomes is evident with increasing passage number (analysis of variance for linear trend p<0.0001). The mean number of human FISH signals per chromosome between passage no. 2 and passage no. 198 increased by a factor of 4.07. (b) A similar exercise was done as above except that mean fluorescent intensity (MFI) of human FISH signals per chromosome, indicative of amplification, was determined. A progressive increase in mean MFI per chromosomes is evident with increasing passage number (analysis of variance for linear trend p<0.0001). The MFI per chromosome increased by a factor of 237.19 between passage no. 2 and passage no. 198. The insets represent partial metaphase images showing human DNA signals on the chromosomes. Blue and red signals represent DAPI and human DNA, respectively.
Figure 7
Concatemers largely comprise open chromatin irrespective of the epigenetic constitution of the host mouse DNA.
Representative images of chromatin fibres immunostained with antibodies against H3k4me3 (red) and H3k9me3 (green). The hybrid concatemers are represented by co-localised yellow signals. The host mouse DNA in the upper image is seen to react with H3k4me3 antibody representing open chromatin, while the host mouse DNA segment in the lower image is comprised of heterochromatin and seen to react with antibody against H3k9me3. However, the concatemers were largely composed of open chromatin irrespective of the epigenetic status of the host mouse DNA. Histograms represent quantitative estimates of H3k4me3, H3k9me3, and hybrid histones after counting 100 fluorescent signals. The values are expressed as mean ± SEM values. Statistical analysis was performed using two-tailed Student’s t-test. * p<0.05, ** p<0.01, and **** p<0.0001.
Figure 8
Concatemers synthesise RNA which is dependent on active cellular metabolism.
Images showing cytoplasmic RNA synthesis by cfChP-treated cells in continuous passage, which is absent in control NIH3T3 cells. Treatment of cfChP-treated cells with Actinomycin D (0.0005 μg/mL) or maintenance at low temperature (31 °C) abolishes RNA synthesis indicating that the latter is dependent on active cellular metabolism.
Figure 9 with 3 supplements
Concatemers synthesise their own protein synthetic machinery.
Chromatin fibres (a) and metaphase spreads (b) were prepared from cfChPs-treated cells in continuous passage. Dual-FISH experiments were performed using a human-specific genomic DNA FISH probe and a probe against human-specific ribosomal RNA, and immune-FISH experiments were performed using a human-specific genomic DNA FISH probe and antibodies against human-specific RNA polymerase III and human-specific ribosomal protein. Co-localised fluorescent signals of DNA and the components of the above protein synthetic machinery are marked with arrows.
Figure 9—figure supplement 1
Control experiments to show that the FISH probes and antibodies used against the components of the protein synthetic machinery viz. ribosomal RNA, RNA polymerase, and ribosomal protein were human specific and did not cross-react with mouse.
(a) Positive control cells using HEK-293 (human embryonic kidney) cells; (b) Negative control experiments using NIH3T3 (mouse embryonic fibroblast) cells.
Figure 9—figure supplement 2
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of human DNA and human ribosomal RNA, human DNA and human RNA polymerase III, and human DNA and human ribosomal protein.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 9—figure supplement 3
Control experiments to detect human DNA signals in control NIH3T3 cells that had not been exposed to cfChPs.
The human-specific FISH probe and antibodies did not detect any signals of human DNA and human ribosomal RNA, human DNA and human RNA polymerase III, and human DNA and human ribosomal protein in mouse cells.
Figure 10 with 3 supplements
Concatemers synthesise a variety of human proteins.
Immuno-FISH analysis on chromatin fibres (a) and metaphase spreads (b) prepared from cfChPs-treated NIH3T3 cells in continuous passage using human specific whole genomic FISH probe and antibodies against various human proteins. The co-localising signals of human DNA and various human proteins are marked with arrows.
Figure 10—figure supplement 1
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of human DNA and various proteins.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 10—figure supplement 2
Control experiments to detect human DNA and protein signals in control NIH3T3 cells that had not been exposed to cfChPs.
The human-specific DNA probe and antibodies did not detect any DNA or protein signals in mouse cells.
Figure 10—figure supplement 3
Control experiments to show that the antibodies used against the various proteins were human specific and did not cross-react with mouse.
Various human and mouse cell lines were probed with human specific antibodies. It can be seen that only human cells reacted with the antibodies and the mouse cells did not.
Figure 11 with 2 supplements
Proteins synthesised by concatemers are fusion proteins.
Dual immunofluorescence analysis using antibody pairs targeting diverse human-specific proteins in chromatin fibres (a) and metaphase spreads (b) prepared from cfChPs-treated NIH3T3 cells in continuous passage. Results show frequent co-localised signals indicative of fusion proteins (arrows). The numbers given in the parenthesis indicate the chromosomal location of the genes that correspond to the proteins.
Figure 11—figure supplement 1
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of different pairs of proteins.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 11—figure supplement 2
Control experiments to show that the antibodies used against the components of fusion proteins that were not included in Figure 10—figure supplement 3 were human specific and did not cross-react with mouse.
A-375 (human melanoma cells); B16F10 (mouse melanoma cells).
Figure 12 with 3 supplements
Concatemers harbour transposable elements.
Dual-FISH analysis on chromatin fibres (a) and metaphase spreads (b) prepared from cfChPs-treated NIH3T3 cells in continuous passage using a human-specific genomic DNA FISH probe and those against human LINE-1 or human Alu show co-localised signals (marked with arrows).
Figure 12—figure supplement 1
Control experiments to show that the probes used against LINE-1 and Alu transposable elements and the antibodies against reverse transcriptase and transposase (please refer to Figure 13 below) were human specific and did not cross-react with mouse.
cfChPs treated cells in continuous passage were used to study species specificity of the probes and antibodies. (a) Positive control cells using A-375 (human melanoma) cells; (b) Negative control experiments using B16F10 (mouse melanoma) cells.
Figure 12—figure supplement 2
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of human DNA and LINE-1, human DNA and Alu.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 12—figure supplement 3
Control experiments to detect human DNA signals in control NIH3T3 cells that had not been exposed to cfChPs.
The human-specific FISH probes did not detect any human DNA or LINE-1 or Alu signals in mouse cells.
Figure 13 with 2 supplements
LINE-1 and Alu sequences are associated with reverse transcriptase and transposase.
Immuno-FISH analysis of chromatin fibres (a) and metaphase spreads (b) prepared from cfChPs-treated NIH3T3 cells in continuous passage using human Alu or LINE-1 probes and antibodies against human reverse transcriptase and human transposase shows co-localising signals (marked with arrows).
Figure 13—figure supplement 1
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of human LINE-1 and human reverse transcriptase, human Alu and human reverse transcriptase, human LINE-1 and human transposase, and human Alu and human transposase.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 13—figure supplement 2
Control experiments to detect human DNA and protein signals in control NIH3T3 cells that had not been exposed to cfChPs.
The human-specific FISH probes and antibodies did not detect any human LINE-1, human Alu, human reverse transcriptase, and human transposase signals in mouse cells.
Figure 14 with 1 supplement
LINE-1 and Alu elements are associated with DNA polymerase and can actively synthesise DNA.
(a) Immuno-FISH images of cfChPs-treated cells in continuous passage showing co-localising signals of LINE-1 (green) and Alu (green) and DNA polymerase γ (orange) and BrdU (red) on chromatin fibres. (b) Similar co-localising signals are shown on metaphase spreads. Co-localised signals are marked with arrows.
Figure 14—figure supplement 1
Histograms representing quantitative results of the degree of co-localisation of fluorescent signals (red and green) of human LINE-1 and human DNA polymerase, human Alu and human DNA polymerase, human LINE-1 and BrdU, and human Alu and BrdU.
Five hundred fluorescent signals were analysed and the number of co-localising red and green signals was estimated. The results were expressed in percentage terms.
Figure 15
LINE-1 and Alu elements proliferate and amplify themselves within the mouse genome over time.
Metaphase spreads were prepared from cells in each progressively increasing passage and probed with a human-specific LINE-1 and Alu probes. (a) The total number of human LINE-1 and Alu signals on the DAPI stained chromosomes was counted, and the mean number of LINE-1 and Alu signals per chromosome was calculated after analysing 15 metaphase spreads at each passage. A progressively increasing LINE-1 and Alu signals per chromosomes is evident with increasing passage number (analysis of variance for linear trend p<0.0001). The number of human LINE-1 signals per chromosome increased by a factor of 7.6 between passage no. 2 and passage no. 198, and by a factor of 6.7 in the case of Alu. (b) A similar exercise was done as above except that mean fluorescent intensity (MFI) of human LINE-1 and Alu per chromosome, indicative of amplification, was determined. A progressively increasing MFI per chromosomes is evident with increasing passage number (analysis of variance for linear trend p<0.0001). Mean MFI per chromosome increased by a factor of 151.3 between passage no. 2 and passage no. 198 in the case of LINE-1, and by a factor of 83.4 in the case of Alu. The insets represent partial metaphase images showing human LINE-1 and Alu signals on the chromosomes. Blue and red signals represent DAPI and human DNA, respectively.
Figure 16
NIH3T3 cells are not unique in their ability to internalise cfChPs.
We used four different cell lines other than NIH3T3 viz. Vero (monkey kidney cells); Dolly (female dog cells); B/CMBA. OV (mouse ovary cells) and HEK293 (human embryonic kidney cells) and tested them for their ability to internalise cfChPs. The cells were treated with cfChPs derived from human serum and chromatin fibres were prepared at 5th passage. The chromatin fibres were probed with a whole genomic DNA probe or with a human specific LINE-1 probe. Human DNA and LINE-1 signal are clearly seen in the treated cells. It is to be noted that the dog FISH probe was specific for X (red) and Y (green) chromosome. Since the cells came from a female dog, only the red X chromosome is seen to react.
Figure 17
The concatemers are largely composed of non-coding DNA.
(a) The human long non-coding RNA probe aligns with human DNA but not with mouse DNA. Chromatin fibres were prepared from two human cell lines viz. HEK293, MRC5 and one mouse cell line viz. NIH3T3 and the cells were hybridised with the long non-coding RNA probe. The results show that the probe has almost 100% coverage in case of the two human cells but does not react with the mouse cells. (b) The concatemers are largely composed of non-coding DNA. cfChPs treated passaged cells were simultaneously probed with a fluorescently labelled human whole genomic DNA FISH probe and the human long non-coding RNA probe (which substituted for a non-coding DNA probe). The degree of co-localisation of fluorescent signals is represented as a histogram, which was generated after counting 1000 human DNA signals, and which shows that the extent of co-localisation of the fluorescent signals was of the order of 98.2%. (c) Confirmation of these findings on metaphase spreads.
Figure 18 with 1 supplement
Biological activities of the concatemers are attributable to non-coding DNA.
Immuno-FISH and dual-FISH analysis of chromatin fibres prepared from cfChPs-treated cells in continuous passage. The chromatin fibres were probed with antibodies against various human proteins and the human non-coding RNA probe (which substituted for a non-coding DNA probe) or the human non-coding RNA probe and FISH probes against LINE-1 and Alu. Co-localisation of various signals is clearly seen. These data suggested that biological activities of the concatemers are attributable to non-coding DNA.
Figure 18—figure supplement 1
Biological activities of the concatemers are attributable to non-coding DNA.
Histograms showing the degree of co-localisation of fluorescent signals of human non-coding DNA or of human whole genomic DNA and those of various biomarkers and LINE-1 and Alu elements. The degree of co-localisation with the biomarkers was similar irrespective of whether a whole human genomic probe or one against non-coding DNA probe was used. These data suggested that biological activities of the concatemers are attributable to non-coding DNA.
Figure 19
NIH3T3 cells treated with cfChPs isolated from the serum of healthy individuals and which were kept in continuous passage show similar characteristics and properties as those of cfChPs isolated from cancer patients.
Figure 20
NIH3T3 cells treated with conditioned medium containing cfChPs released from dying MDA-MB-231 breast cancer cells and which were kept in continuous passage show similar characteristics and properties as those of cfChPs isolated from cancer patients and healthy individuals.
Figure 21
Representative image to confirm that fibres prepared from NIH3T3 cells are chromatin fibres and not DNA fibres.
The fibres were stained with DAPI and anti-histone H4 antibody.
Videos
Video 1
A complete summary of our work.
Additional files
-
MDAR checklist
- https://cdn.elifesciences.org/articles/103771/elife-103771-mdarchecklist1-v1.docx
-
Supplementary file 1
Antibodies and FISH probes used in this study (custom synthesized as per vendors’ specifications).
- https://cdn.elifesciences.org/articles/103771/elife-103771-supp1-v1.docx
-
Supplementary file 2
Clinical and demographic information of cancer patients and healthy individuals who provided blood samples for isolation of cell-free chromatin particles.
- https://cdn.elifesciences.org/articles/103771/elife-103771-supp2-v1.docx
-
Supplementary file 3
Cell lines used in this study.
- https://cdn.elifesciences.org/articles/103771/elife-103771-supp3-v1.docx
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Horizontally transferred cell-free chromatin particles function as autonomous ‘satellite genomes’ and vehicles for transposable elements within host cells
eLife 13:RP103771.
https://doi.org/10.7554/eLife.103771.3
